Re: include statement bugs
Jason Mirwald wrote:
----------
> From: Jason Mirwald <guest at RapidEuphoria.com>
> To: EUforum at topica.com
> Subject: Re: include statement bugs
> Sent: 19 oct 2004 y. 6:59
>
> posted by: Jason Mirwald <jmirwald at ameritech.net>
>
> This is absolutely ridiculous in my opinion. In all the time that people
> have invested in beating this dead horse, why hasn't anyone seriously
> discussed the idea that we could all use one set of include files?
I think, there is no the task for too big set of the existing
include files. Say, just win32lib + w32engine have about 10000
global names at all.
You can program any thing you want just now.
Most of users just do not need new sets of libs.
Include and use existing ones.
win32lib + w32engine + bilingual Euphoria with the unlimited
alphabet has about 11000 global names at all.
Almost all Windows API functions + power of Euphoria itself
are under your fingers just now.
win32lib includes the standard EU libs.
The standard EU libs - just about 20 files, about 160K at all.
The rest tools and programs are created by happy users - more
than 1300 packages, about 400 authors.
If you have some good idea for programming, program please,
all tools are ready for you just now and foreve.
> Obviously more of the archives are going to be broken no matter what
> happens, and all my code will be crippled, as will the code of others I
> tend to share with, but it isn't like the entire set couldn't be replaced
> at least 10 times over.
>
> It's painfully obvious that the includes distributed with the interpreter
> are inadequate. They've been virtually unchanged for at least 5 years (no
> major changes), and Euphoria is still stuck right where it was at least 5
> years ago. This isn't just a coincidence, this is the REASON for the lack
> of advancement.
There is no any need to distribute by RDS itself something additional
to RDS standard package - download the needed package from EU sites
just now and use. RDS has the largest of EU sites.
There is no any need to co-ordinate the activity of those
independent authors. They work and contribute, work and contribute.
What else?
Then, include files by RDS allow you to *remake* all Archive,
all 1300 packages, if you want, from the scratch.
This is the real power of that simple little language.
> First of all, Rob has never expressed any support for this idea that I
> have ever seen or heard of. This lends to the idea that he only endorses
> the files that are installed with the interpreter, and in turn causes
> most people to build upon those includes. Since they are just not enough
> in most cases, almost everyone who uses them must write their own
> includes to expand upon that code, and this in turn leads to conflicts
> with other libraries, because EVERYONE is doing it.
It is a good thing - the end users can do all they need themselves,
and do it themselves and do it one for others.
> Then, the community in general seems to either oppose, or hold absolutely
> no interest in, any set of files that can't replace the original set with
> ABSOLUTELY NO PROBLEMS OR MODIFICATIONS.
You must to prove that the original set requires to be replaced
with something else. Prove first of all.
I do not like if someone returns from the vacation and says -
all around is shit, crap, shitty crap, crappy shit, I'll replace
all that just now with my own big one, but only under yours name.
Sorry please, life is life, I do not like these words, but they
are the most suitable now, I think.
> So now the problem is finally
> forced upon us? I, for one, don't think so. It seems to me that the
> community has DECIDED that Euphoria should remain right where it's at.
>
> If this language is going to advance and grow, everyone needs to get past
> the notion that we can keep backward compatibility with everything ever
> written for ever and ever. With the proper planning and support from more
> than about three people, we could work toward a better, more
> comprehensive set of includes, that would alleviate ALL these problems in
> one fell swoop. The road won't be easy, but am I the only one who can see
> that if an include file is worth re-using, modifying it ONE TIME to run
> on a new, and better set, is still easier than writing a different
> version for each user who wants to include it?
>
> Jason
Oh-ho-ho-oh, Jason, let us wait for 2.5 and test it first of all. Ok?
Regards,
Igor Kachan
kinz at peterlink.ru
|
Not Categorized, Please Help
|
|
